Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 3 - Tests for Variability in Methodology

 
 
Subscribe to our free E-Newsletter, "Agri-News" (formerly RTW This Week)Agri-News
This Week
 
 
 
 Comparison of standard readings | Settling time | Light | Temperature | Accuracy of measuring scoop | Drying times

Comparison of Standard Readings in Morning and Afternoon

There was a statistically significant difference between the readings of the 0.005M standard from the morning to the afternoon (Table 1). Therefore, in order to decrease the variability that might result from preparing and reading the standards once a day we decided to prepare the standard solutions twice a day, in the morning and afternoon. We then applied the resulting standard curve to those samples tested at the corresponding time period, which allowed us to test a greater quantity of samples per day. This differed from the procedure followed by Weil et al. (2003) who completed readings of the standards for each batch of three samples.

Table 1. The Descriptive Statistics and Results of Paired T-test for Standard Readings.

*significant difference at p<0.05 between morning and afternoon readings.

This was based on five readings, in both the morning and afternoon, for the 0.005, 0.01 and 0.05M standards and three readings for the 0.015M standard.

Settling Time

To test the effects of settling time on the readings we used 12 soil specimens from samples taken during fall 2002 Soil Quality benchmark sampling, which varied across textural classes and management practices. We prepared the samples according to the protocol followed by Weil et al. (2003) and obtained the absorbance readings after five time periods, 10, 15, 20, 30 and 60 minutes, had elapsed. We found there to be a statistically significant difference in the readings taken at 10 minutes to those taken at 20, 30 and 60 minutes (Table 2). Therefore, in order to maximize the number of samples that could be tested at one time and to keep variability as low as possible, we decided that no more than 10 samples could be analyzed at once to ensure none exceeded a 17 minute settling time.

Table 2. Absorbance Readings and Results of Paired T-tests on 2002 Soils Across Varying Textural Classes.

*significant difference at p<0.05 when compared to readings taken.

Light

Exposure of the KMnO4 solution to light can increase the rate of decomposition (Weil et al. 2003). We tested the readings of standard solutions at various locations of differing light exposure at Bonaventure and found no statistically significant differences between locations (Table 3).

Table 3. Readings of Standards at Various Locations.

Date 1/14/2004
.
Readings of standards
Means
Location
Rep Number
0.005M
0.01M
0.02M
0.005M
0.01M
0.02M
Coolers
1
0.24
0.49
0.84
0.26
0.51
0.81
.
2
0.26
0.51
0.76
.
.
.
.
3
0.27
0.53
0.83
.
.
.
Grinders
1
0.25
0.52
0.86
0.25
0.54
0.80
.
2
0.26
0.53
0.81
.
.
.
.
3
0.24
0.57
0.73
.
.
.
Scale bench
1
0.24
0.46
0.9
0.24
0.50
0.87
.
2
0.23
0.52
0.85
.
.
.
.
3
0.24
0.53
0.85
.
.
.
In direct sun
1
0.24
0.46
0.96
0.24
0.49
0.91
.
2
0.25
0.51
0.93
.
.
.
.
3
0.24
0.49
0.85
.
.
.
Range of readings
.
0.23-0.27
0.46-0.57
0.73-0.96
.
.
.

Temperature

We recorded the temperature of the shop once in the morning and again in the afternoon. We found that it was generally cooler in the morning than the afternoon. The average ambient temperature was 20.17 ºC and the temperature ranged by 1.8 °C.

Accuracy of Measuring Scoop

The protocol followed by Weil et al. (2003) uses a scoop to measure out the five grams of soil required for testing. To assess the accuracy of our measuring scoop to this requirement we weighed 11 soil samples with varying soil properties using a metal five-milliliter scoop. We ran the test according to the following procedure:

1)Take the scoop and place on scale, tare scale.
2)Take scoop of dry soil and level.
3)Place scoop with soil on scale and weigh, record the weight of the soil.
4)Take soil that was in scoop and place in separate container.
5)Repeat steps 1-4 four more times for the sample.
6)After one sample has been weighed 5 times clean off the scoop and re-weigh the scoop and tare it again if needed.
7)Proceed to the other samples repeating steps 1-6.

We found that the scoop weights of the soils ranged from 3.08 grams to 5.07 grams and the mean weight of all the soil measured using the scoop was 3.91 grams (Table 4). In order to decrease the variability caused by the inaccuracy of the scoop we substituted the value of 0.005 kg with 0.00391 kg into the equation for calculating the active C (mg/kg) (Appendix 4).

Table 4. Results of Measuring Accuracy Test for Weil et al. (2003) Procedure.
Rep Number
Mean (g)
Range (g)
Variance
Standard Deviation
Coefficient of Variation
Sample I.D
1
2
3
4
5
DAPP
3.22
3.67
3.48
3.63
4.14
3.63
0.92
0.11
0.34
9.27
Carstairs
3.08
3.14
3.24
3.28
3.22
3.19
0.2
0.01
0.08
2.53
Youngstown
5.02
4.85
4.89
5.07
4.75
4.92
0.32
0.02
0.13
2.64
Tilley
4.71
4.45
4.4
4.25
4.57
4.48
0.46
0.03
0.17
3.89
Enchant
3.83
3.83
4.17
4.1
3.78
3.94
0.39
0.03
0.18
4.54
Boyle
3.43
3.52
3.32
3.43
3.19
3.38
0.33
0.02
0.13
3.75
Beiseker
3.57
4.11
4.11
3.96
4.34
4.02
0.77
0.08
0.28
7.09
Dunmore
3.56
3.7
4.07
3.95
4.46
3.95
0.9
0.12
0.35
8.86
Warspite
4.02
3.99
4.22
3.89
3.56
3.94
0.46
0.06
0.24
6.15
Lacombe
3.41
3.5
3.59
3.62
4.14
3.65
0.73
0.08
0.28
7.80
Etzicom
4.21
4.25
4.2
3.37
3.76
3.96
0.88
0.15
0.38
9.72
Overall Mean . .. . .
3.91
.. . .

Drying Times

Our study utilized 41 field moist samples taken from AESA Soil Quality benchmark sites across Alberta. We decided to test the effects of air-drying all the samples for a period of 24 hours (or until all samples were of equal dryness) as compared to drying the field moist samples for the 15-minute drying time suggested by Weil et al. (2003).

A sub-sample (5 scoops) of each of the 11 moist soil samples was taken and crumbled gently and allowed to dry in an aluminum tray. The sample was evenly spread out in the tray and left to dry until all samples were of equal dryness. The samples were then tested according to the procedure outlined in Appendix 4 and replicated three times.

Meanwhile, another sub-sample of 5 scoops was taken from the 11 samples. Each of these was crumbled gently and laid out to dry on a black rubber mat. The samples were spread out evenly and thinly and placed in direct sunlight when possible. The soils were allowed to dry for 15 minutes and were mixed three times to help ensure that they dried as evenly as possible. When the drying time was up the samples were tested based on the procedures in Appendix 4.2 and replicated three times.

We found that in four of the 11 sites there was a statistically significant difference between soils dried for 15 minutes and those dried for 24 hours (Table 5). Higher standard deviations, standard errors and variances were recorded for soils dried for 15 minutes, with the exception of one site. In order to minimize this variability we decided to dry the soils for 24 hours.

Table 5. Readings and Descriptive Statistics of 24 hr and 15-Minute Dried Soils and the Results of a T-test.

Readings
15 Minute Drying Time
24 Hour Drying Time
Site
Name
 15 min
 24 hrs
Mean
Standard
Error
Standard
Deviation
Variance
Mean
Standard
Error
Standard
Deviation
Variance
Dapp
0.46
0.28
0.48
0.0841
0.1457
0.0212
0.32
0.0203
0.0351
0.0012
.
0.34
0.35
.
.
.
.
.
.
.
.
.
0.63
0.32
.
.
.
.
.
.
.
.
* Carstairs
0.32
0.08
0.31
0.0291
0.0503
0.0025
0.07
0.0186
0.0321
0.0010
.
0.26
0.09
.
.
.
.
.
.
.
.
.
0.36
0.03
.
.
.
.
.
.
.
.
Youngstown
0.51
0.52
0.58
0.0636
0.1102
0.0121
0.57
0.0404
0.0700
0.0049
.
0.53
0.54
.
.
.
.
.
.
..
.
.
0.71
0.65
.
.
.
.
.
.
.
.
Tilley
0.63
0.57
0.62
0.0186
0.0321
0.0010
0.67
0.0491
0.0850
0.0072
0.58
0.7
0.64
0.73
* Enchant
0.61
0.45
0.65
0.0633
0.1097
0.0120
0.44
0.0296
0.0513
0.0026
.
0.56
0.48
.
.
.
.
.
.
.
.
.
0.77
0.38
.
.
.
.
.
.
.
.
* Boyle
0.48
0.36
0.52
0.0536
0.0929
0.0086
0.34
0.0252
0.0436
0.0019
.
0.46
0.37
.
.
.
.
.
.
.
.
.
0.63
0.29
.
.
.
.
.
.
.
.
Beiseker
0.37
0.33
0.47
0.0612
0.1060
0.0112
0.33
0.0376
0.0651
0.0042
.
0.45
0.39
.
.
.
.
.
.
.
.
.
0.58
0.26
.
.
.
.
.
.
.
.
Dunmore
0.54
0.52
0.68
0.0872
0.1510
0.0228
0.57
0.0677
0.1172
0.0137
.
0.66
0.7
.
.
.
.
.
.
.
.
.
0.84
0.48
.
.
.
.
.
.
.
.
Warspite
0.38
0.41
0.55
0.1041
0.1804
0.0325
0.42
0.0406
0.0702
0.0049
.
0.54
0.49
.
.
.
.
.
.
.
.
.
0.74
0.35
.
.
.
.
.
.
.
.
*Lacombe
0.24
0.14
0.28
0.0219
0.0379
0.0014
0.13
0.0176
0.0306
0.0009
0.31
0.16
0.3
0.1
Etzikom-Irrig
0.44
0.47
0.59
0.0801
0.1387
0.0192
0.47
0.0549
0.0950
0.0090
0.63
0.57
0.71
0.38
* indicates a statistically significant difference between the 15 minute and 24 hour dried soils at p<0.05.
 
 
 
 

Other Documents in the Series

 
  Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Abstract
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Introduction
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Results and Discussion
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Conclusions and References
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 1 - Methodologies for Field Tests
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 2 - Lab Procedures for LFC and OM (%)
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 3 - Tests for Variability in Methodology - Current Document
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 4 - Procedure for Active C Test
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 5 - Procedure for Basic EDTA Test
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 6 - Descriptive Statistics for Results of Active C Methods
Evaluation of Two Field Methods to Estimate Soil Organic Matter In Alberta Soils: Appendix 7 - Descriptive Statistics for Results of Basic EDTA Method
 
 
 
 
Share via AddThis.com
For more information about the content of this document, contact Len Kryzanowski.
This document is maintained by Laura Thygesen.
This information published to the web on May 4, 2004.
Last Reviewed/Revised on November 6, 2018.